HGPU group<p>Demystifying NCCL: An In-depth Analysis of GPU Communication Protocols and Algorithms</p><p><a href="https://mast.hpc.social/tags/CUDA" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>CUDA</span></a> <a href="https://mast.hpc.social/tags/GPUcluster" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>GPUcluster</span></a> <a href="https://mast.hpc.social/tags/Communication" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Communication</span></a></p><p><a href="https://hgpu.org/?p=30035" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://</span><span class="">hgpu.org/?p=30035</span><span class="invisible"></span></a></p>